Large Memory
These job classes are intended for general purpose workloads, with an emphasis on high throughput for large numbers of relatively small jobs. All jobs should be submitted with a #PBS resource request of the form:
- #PBS -lselect=1:ncpus=X:ompthreads=X:mem=Ygb
- #PBS -lwalltime=HH:MM:00
Where X and Y are the number of cores and amount of memory respectively.
The following job geometries are supported:
| Job class | Number of nodes | ncpus | Max mem | Max walltime/hr | Max number of concurrently running jobs per user |
| largemem | 1 | 10n | 120n GB | 48hr | 50 |
where n is a multiplier between 1 and 100, inclusive. The same multiple must be used for both ncpus and mem.
Hybrid jobs – those using a mix of MPI and OpenMP – are permitted but strongly discouraged in the largemem class.
Jobs found to be using markedly less memory than requested are liable to be terminated.
Estimating job start times
The command qstat -w -T can be used to show projected start times for jobs. These are estimates, not guarantees. Use them for guidance but do not rely on them. They may in some cases be quite inaccurate.